Ancient Narrative Volume 2

The Satyrica of Petronius as a Roman Palimpsest

GOTTSKÁLK JENSSON Reykjavík

Roman literature is possibly the most palimpsestuous of literatures.1 So much so that lovers of Roman letters have had to fight off the unwanted comparison with Roman plastic arts where, as is well known, there are no originals. A series of famous names will emphasise the obvious: Plautus, Terence, Varro, Cicero, Lucretius, Catullus, Virgil, Livy, Horace, Ovid, Seneca, Statius and Apuleius. These (and many more) were borrowers, reworkers, rewriters. But we try not to hold it against them. When the stuff of Roman literature is Greek (and often, too, when it is Roman) the method of writing is translation, transformation, and imitation. By the same process that Roman literature grew out of Greek literature, much of Western literature grew out of Latin literature. In fact, ever since the Romantics, we have been extremely reluctant to admit to this influence, to any influence in fact, instead fantasizing about ‘originality’, or what one theorist sees as a longing for a freedom from paternal influence.2 Of course, demanding originality from Roman literature is to apply later esthetics to earlier art, but the anachronism doesn’t usually bother Petronian scholars. Few Roman writers have been more fantasized about in this manner than the elusive author of the Satyrica. We shall see in the central part of this paper that the modern ‘Petronius’ was invented in the late 19th century under ideological pressure. Although we rarely admit to this, we know nothing about ————— 1 I use the term ‘palimpsest’ as a broad term to denote a text derived from a previous text through transformation or imitation. It covers translation, copy, make-over, adaptation and many other such terms. For the purpose of this article the broad sense of the word is defined by Gérard Genette’s Palimpsests: Literature in the Second Degree. Translated by Channa Newman and Claude Doubinsky (Lincoln and London 1997; original French publication 1982). According to Genette (399) the adjective ‘palimpsestuous’ was coined by Philippe Lejeune. 2 For this Romantic and Post-Romantic psychology of literature, see Harold Bloom’s classic The Anxiety of Influence: A Theory of Poetry (New York, 1973). THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 87 the author with certainty.3 If we ignore the usual guesswork, too uncertain in any case to provide a substantial foundation, we are left with no more than a name on the title page. And yet so much meaning has been invested in that name, so much affection has been bestowed since the 19th century upon this putatively ‘original’ Roman novelist, that the mere proposal that the Satyrica is a typical Roman palimpsest evokes a sense that a line of propriety has been crossed. Not surprisingly, the proposal has never been made before. One Scottish scholar,4 it is true, in an influential study on the ‘Roman Novel’, ridicules the Frenchman Collignon and the American Perry because they ‘regard the Satyricon as the adaptation of a lost romance perhaps called Priapeia.’ But he misrepresents the writings of these scholars. Albert Col- lignon, too, believed Petronius to be an original author and merely suggested that he might perhaps have used some plot ideas from a Greek model.5 Ben Edwin Perry was likewise a firm believer in Petronian originality and merely claimed that he used the same method as the author of the Greek Ass Story and wrote the central fable ‘on the basis of folklore plots’.6 The Oxford profes- sor Peter Parsons certainly did write the notorious words: ‘Natural reason long ago revealed that Petronius had a Greek model,’ but he didn’t mean a specific Greek work that Petronius had adapted, but a genre that Petronius had imitated.7 As much is revealed by his reference to ‘the Greek Schelmenroman’ and the fact that Parsons qualifies his statement with a footnote reference to a certain German publication from the early 20th century.8 If we make the effort to follow up on the reference, we find that Ulrich von Wilamowitz- Möllendorff in no weak terms affirms Petronian originality (Dem Dichter soll wahrlich seine Originalität nicht verkleinert werden), and merely speculates as to whether there existed a Greek roman comique or Schelmenroman, subse- quently granting with an evident grudge that ‘das picarische Element’ in Petronius is as Greek as the wrath of Priapus, the Widow of Ephesos and the ————— 3 Rose, 1971, persuaded many with his arguments for the identification of the author with C. Petronius, described as Nero’s elegantiarum arbiter by Tacitus (Ann. 16.18–19), but the identification is plausible at best and the case remains inconclusive. Smith 1975, xii– xiv, 213–214, has pointed out some of the weaknesses of this identification. 4 Walsh 1970, 17 5 Collignon 1892, 323. 6 Perry 1925, 39f. 7 Parsons 1971, 66. As to the notoriety of these words, cf. Stephens and Winkler 1995, 364–366. 8 The footnote runs like this: ‘e.g. Die Kultur der Gegenwart I viii, Die Gr. u. Lat. Litera- tur (3rd edn 1912) 190 (Willamowitz), 459 (Leo).’ 88 GOTTSKÁLK JENSSON shipwreck. At the end of this concession he adds the words: ‘... as is taught by the obvious and confirmed by analysis’ (lehrt der Augenschein und bestätigt die Analyse), which is clearly the phrase Parsons has anglicized by his ‘natural reason ... revealed’. We find nothing here or elsewhere regarding a specific Greek text that Petronius may have adapted. What is meant by a ‘Greek model’ in Petronian scholarship is always either a ‘serious’ Greek novel to be parodied, or some Greek genre which is designated by some such label as ‘comic’, ‘criminal’ or the anachronistic ‘picaresque’ (from the Spanish word picaro) and its German translation Schelmenroman. The possibility of a Greek text, the Σατυρικά, directly adopted by Petronius has never been entertained before, not even when scholars have attempted to list all the hypothetical possibilities.9 So, either it is simply a foolish idea, a philological absurdity, or it has become one of those things that you better not suggest, if you care for your respectability in the scholarly community: a disciplinary taboo. At the risk of making an ass of myself, I shall now try to show that, philo- logically, the suggestion is not unsound. What we are dealing with is a Roman narrative with a Greek title, Satyrica (Σατυρικά),10 told by a Greek called En- colpius (Εγκόλπιος)—hence the Greek accusative ‘Encolpion’ (Sat. 92,7; 104,1; 109,3; 114,9; 128,7)—from the Greek city of Massalia (Μασσαλία). In short a Greek story populated by Greek characters moving in a Greek environment. Given this general Greekness of the story, is it not possible that there existed a Greek novel by Encolpius of Massalia (almost certainly a pseudo- nym) relating the travels of an exiled anti-Odysseus? Can we rule out palimpsest because the Greek text is never even mentioned in extant Greco-Roman

————— 9 Stephens and Winkler 1995, 364–365 n. 17, do just this when they ask in continuation of Parsons’ dictum: ‘what kind of Greek model? Greek novels of the historical or “idealistic” type, as well as salacious stories like Aristides’ Milesiaka certainly preexisted the Sa- tyrica, and Petronius, educated Roman that he was, would surely have read what existed. Did he adapt or satirize what had come to be a generic plot, did he have an individual serious novel in mind when he wrote the Satyrica, or was he writing a style of criminal- satiric fiction already well established in Greek?’ 10 The title of the earliest and best MS is Satiricon, or Satyricon. The fourth century writer Marius Victorinus also preserves Satyricon (GL 6, 153). The Latin spelling Satyricon (sc. libri) stands for the Greek genitive plural Σατυρικῶν which gives Σατυρικά in the nominative. Henriksson 1956, 77, concludes in his study of Greek book-titles in Roman literature that the Roman readership of Petronius could probably not differentiate the meaning of the forms Satiricon and Satyricon, since there is no indication that such etymological understanding existed. In other words, satyrs and satire were related concepts. THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 89 literature? Despite the lack of any remains of a Greek story, the problem of the Greekness of the Satyrica will not go away, because such a multitude of texts is completely lost and Roman literary history teaches us that Greek material in Roman literature originates from specific sources. As we shall see later in the article, at a certain point in the reception of the text it was indeed necessary to explain away the Greekness of the Satyrica before it could be read as a quintessential ‘national’ Roman novel of the realistic type. If one argues from the premise of ancient literary history alone, there is nothing improbable in the hypothesis that an ancient Roman writer composed a Latin fictional narrative by loosely rewriting a Greek text, yet retaining both the Greek title and the name of the Greek narrator, along with much of the rest of the story. We can say with certainty that there is nothing improbable in this hypothesis, because this is what Cornelius Sisenna did in the first century B.C., when he created his Milesiae or Milesiarum libri, a Latin version of the lost Greek Milesiaca (Μιλησιακά) by Aristides of Miletus;11 and because this is what Apuleius did, in the second century, when he produced the Metamor- phoses, a Latin version of the partly preserved Greek Metamorphoseis (Μεταµορφώσεις) of Lucius of Patras. There is no need to mention other Ro- man genres built on Greek works. Nothing in the ancient testimony contradicts this possibility. The Lucianic dialogue Erotes shows that the Greek Milesiaca (Μιλησιακά) had the form of recollections, narrated by the central narrative persona of Aristides, who every now and then related how he encountered other people who told him stories, which he then incorporated into his own novelistic narrative by retelling them in the persona of the individual from whom he claimed to have heard them, thus playfully creating a distance between himself and the obscene material he related.12 Ovid’s Tristia and Plutarch’s Life of Crassus support the testimony of the Lucianic dialogue, and add that a Roman, Cornelius Sisenna, created a Latin version of the Milesiaca (Μιλησιακά) in the first century B.C. Apuleius himself tells us in the opening of his Latin adaptation of the Metamorphoseis (Μεταµορφώσεις) that the story is told in the sermo Milesius, the Milesian discourse. Since the Byzantine scholar Photios describes the Metamorphoseis (Μεταµορφώσεις) as being in several volumes and ‘stuffed with fabulous

————— 11 For the evidence of the title, see Harrison 1998. 12 This structure was first revealed by Bürger 1892, whose thesis has now been restated by Harrison 1998, who thoroughly examines the testimony about Milesian tales, and myself in a Ph.D. dissertation from 1996, 304–319 (forthcoming in print as AN Supplementum). 90 GOTTSKÁLK JENSSON stories and shameless obscenity’ (Phot. Bibl. Cod. 129, γέµει ... πλασµάτων µὲν µυθικῶν, ἀρρητοποιΐας δὲ αἰσχρᾶς) there is little reason to doubt that the Greek story had the same narrative structure as its Latin version, although this structure is only partly preserved in the epitome Lukios or the Ass (Λούκιος ἣ Ὄνος). As for the similarity of these texts with the Satyrica, Macrobius tells us, in a commentary on Cicero’s Dream of Scipio, that Petronius and Apuleius wrote the same kind of stories (Somn. 1.2.7–8). We can furthermore ascertain from the distinct and unusual narrative form of the Satyrica that it is written in the sermo Milesius, i.e. it has the same narrative structure (described by Lucian in his Erotes) as the Milesiaca, and the Metamorphoses of Apuleius. The Apuleian term sermo Milesius implies both a generic relationship with the Μιλησιακά and Roman palimpsest, since the reference in Apuleius’ Latin must be to Sisenna’s Latin adaptation of that work. It is thus Apuleius’ generic label for his palimpsest, and probably wasn’t in the Metamorphoseis. According to its ancient description, this type of narration should be thought of as imitating oral and performative story-telling, i.e. it should be analyzed and described with reference to the paradigm of a single story-teller, a single speaking voice, capable of impersonating the characters of the story. In the following model of the narrative speaker-personae in the Satyrica the proper names may be thought of as verbal equivalents of masks (the narrator is marked by caps, ENCOLPIUS, even in minimal bridges crossing from one impersonation to another, but the impersonated masks by quotation marks, e.g., ‘Agamemnon’). By accident of preservation, the extant Satyrica opens in the middle of a passage where the narrator is impersonating his younger self, a character in the story, as he spoke at that moment in the past, after which the central identity resurfaces and so on and so forth:

[...] ‘Encolpius’ — ENCOLPIUS (3.1) — ‘Agamemnon’ — ENCOL- PIUS (6.1–17.3) — ‘Quartilla’ — ENCOLPIUS (18.1–37.1) — ‘Her- meros’ — ENCOLPIUS (39.1) — ‘Trimalchio’ — ENCOLPIUS (40.1– 42.1) — ‘Seleucus’ — ENCOLPIUS (43.1) — ‘Phileros’ — ENCOL- PIUS (44.1) — ‘Ganymedes’ — ENCOLPIUS (45.1) — ‘Echion’ — ENCOLPIUS (47.1) — ‘Trimalchio’ — ENCOLPIUS (47.7–50.4) — ‘Trimalchio’ — ENCOLPIUS (52.4–55.4) — ‘Trimalchio’ — ENCOL- PIUS (56.7–57.1) — ‘Hermeros’ — ENCOLPIUS (59.1–61.5) — ‘Ni- ceros’ — ENCOLPIUS (63.1) — ‘Trimalchio’ — ENCOLPIUS (64.1– THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 91 (64.1–65.9) — ‘Habinnas’ — ENCOLPIUS (67.2–71.5) — ‘Trimalchio’ — ENCOLPIUS (72.1–74.13) — ‘Trimalchio’ — ENCOLPIUS (78.1– 81.2) — ‘Encolpius’ — ENCOLPIUS (82.1–83.7) — ‘Eumolpus’ — ENCOLPIUS (90.1–101.9) — ‘Eumolpus’ — ENCOLPIUS (102.10) — ‘Encolpius’ — ENCOLPIUS (102.14) — ‘Giton’ — ENCOLPIUS (103.3–106.4) — ‘Eumolpus’ — ENCOLPIUS (107.7) — ‘Lichas’ — ENCOLPIUS (107.12–108.14) — ‘Tryphaena’ — ENCOLPIUS (109.1– 110.8) — ‘Eumolpus’ — ENCOLPIUS (113.1–115.9) — ‘Encolpius’ — ENCOLPIUS (115.20–116.3) — ‘vilicus quisdam’ — ENCOLPIUS (117.1–13) — ‘Eumolpus’ — ENCOLPIUS (124.2–125.4) — ‘Chrysis’ — ENCOLPIUS (126.8–129.3) — ‘Circe’ — ENCOLPIUS (129.10–12) — ‘Encolpius’ — ENCOLPIUS (130.7–132.9) — ‘Encolpius’ — EN- COLPIUS (132.11–12) — ‘Encolpius’ — ENCOLPIUS (133.1–2) — ‘Encolpius’ — ENCOLPIUS (133.4–134.10) — ‘Oenothea’ — EN- COLPIUS (135.1–138.5) — ‘Encolpius’ — ENCOLPIUS (139.3–141.1) — ‘Eumolpus’ […].

A similar model of the narrative speaker-personae in the Metamorphoses of Apuleius would look like this:

LUCIUS (1.1–5) — ‘Aristomenes’ — LUCIUS (1.21–2.13) — ‘Milo’ — LUCIUS (2.15–21) — ‘Thelyphron’ — LUCIUS (2.31–3.3) — ‘accusa- tor quidam’ — LUCIUS (3.4) — ‘Lucius’ — LUCIUS (3.7–15) — ‘Fotis’ — LUCIUS (3.19–4.9) — ‘unus ex numero latronum’ — LUCIUS (4.22–26) — ‘Charite’ — LUCIUS (4.27) — ‘delira et te- mulenta illa anicula’ — LUCIUS (6.25–28) — ‘Charite’ — LUCIUS (6.30–31) — ‘unus latronum’ — LUCIUS (6.32–7.1) — ‘quidam de numero latronum’ — LUCIUS (7.2–5) — ‘Haemus (sive Tlepolemus)’ — LUCIUS (7.10–8.1) — ‘unus ex famulis Charites’ — LUCIUS (8.15– 20) — ‘senex magnus’ — LUCIUS (8.21–9.16) — ‘anicula’ — LUCIUS (9.22–24) — ‘maritus’ — LUCIUS (9.26–8) — ‘medicus’ — LUCIUS (10.12–11.2) — ‘Lucius’ — LUCIUS (11.3–4) — ‘Isis’ — LUCIUS (11.7–15) — ‘sacerdos’ — LUCIUS (11.15–25) — ‘Lucius’ — LUCIUS (11.26–29) — ‘imago’ — LUCIUS (11.30).

Note that the characters and names in both stories are mostly Greek. Behind the Roman name Lucius is the Greek Λούκιος. One would be hard pressed to 92 GOTTSKÁLK JENSSON find a similar use of Greek characters and names in Roman literature, except in Latin New Comedy, Greco-Roman Tragedy or Greco-Roman Epic on Greek subjects, i.e. Latin palimpsests written over Greek texts, where the common practice was to keep the original names.13 As for the relationship of the Greek narrators to the authors of these works, we know that Lucius of Patras was the narrator of the Greek Metamorphoseis (Μεταµορφώσεις), just as Aristides of Miletus narrated the Milesiaca (Μιλησιακά). Loukios of Patras is generally thought to be a fictional identity, but the historical identity of Aristides is no less obscure. Encolpius of Massalia is therefore no more and no less historical than the others. All three Greek works could have been anonymous. In the context of the Milesiaca, Ovid adds a third anonymous author who ‘recently composed the Sybaritica’ (Trist. 2.417, nec qui composuit nuper Sybaritica), which seems to be the same text as the Sybaritici libelli referred to by Martial (12.95,1–2) as the emulated Greek model of a pornographic composition in Latin from the stylus of a certain Mussetius (Musseti pathicissimos libellos, / Qui certant Sybariticis libel- lis).

The lack of cultural and linguistic realism in Roman poetic palimpsests has been studied by Gordon Williams. In a truly insightful chapter, ‘The blending of Greek and Roman’, Williams explains how Roman authors acted as if the transition from Greek to Roman literature was a natural continuation of the same tradition: ‘Roman poets treated both earlier Roman poets and Greek poets in the same way that Greek poets had themselves treated their own predecessors’.14 This is only surprising to us, because we are accustomed to assigning a ‘nationality’ to literary works and explaining them as French, German, English etc. A Roman palimpsest is neither a translation, which pre- supposes that one language can function as the unproblematic parallel of another, nor a complete reworking, which transforms cultural settings and forces them to comply with the new environment. Instead, Roman palimpsests blend Greek and Roman elements in such an undifferentiated manner that the final outcome is, from our point of view, a utopian creation. ————— 13 The Greek names of the puellae or boys of the poems of e.g. Catullus, Tibullus, Proper- tius and Ovid do not really comprise a good counterexample, for they have a special function as pseudonyms (cf. Apuleius Apol. 10,5). Nor is the Greekness of these names completely meaningless. Catullus for one uses the name ‘Lesbia’ to intimate his debt to the poetry of Sappho. 14 Williams 1968, 254. THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 93 These symptoms are obvious in those works which we know to be direct Roman adaptations from Greek literature, such as the works Williams makes the objects of his study, the comedies of Plautus, Virgil’s Eclogues, and Horace’s Odes. Despite the process of transformation, which adds much Ro- man material in the Latin language so that a peculiar Greco-Roman cultural amalgam is created which nowhere existed in antiquity, the Satyrica has a cultural milieu that is recognizably Greek and based on a ‘tradition of mali- cious erotic ethnography’15 and the mythologized identities of Greek cities. The setting for the last episode of the extant Satyrica is the city of Croton. Since this is the only intact introduction to a Greek city in the story as we have it, it is all the more significant. Why Croton? What wars are being referred to in the introduction of Cro- ton as a city which has ‘squandered its wealth in frequent wars’ (Sat. 116, post attritas bellis frequentibus opes)? The ancient Greek colony of Croton (Κρότων) is best known in literature for destroying great and luxurious Sybaris in 510 B.C. Sybaris had been closely affiliated with wealthy and powerful Miletos which, in turn, saw its golden age end in the late fifth century. Like luxurious Sybaris and wealthy Miletos, powerful Croton is a legend of the distant past, from the period after the Greek colonial expansion. Such tales of the life in famous ancient cities may have been termed ‘community legends’ µῦθοι πολιτικοί (Schol. Arist. Vesp. 1259a), and are worthy of consideration as a genre on their own. Sybaris was proverbial for the same quality that made the Milesians notorious, namely luxury and licentious behavior. The city of Croton in the Satyrica, a place without regenerative power that feeds on the old and has made legacy hunting into a way of life, is therefore not a realistic description of Roman Italy, but a fantasy and a legend of the distant past. The choice of Croton in the Satyrica obeys the same logic as the choice of Thes- saly—legendary for witchcraft—for Lucius’ encounter with witches in the Greek Metamorphoses (Μεταµορφώσεις), and the choice of luxurious Miletos as the setting for the licentious and obscene stories told in the Milesiaca (Μιλησιακά) of Aristides. Croton, however, is not the city that defines the subject of the Satyrica in the same manner that Thessaly defines the subject of the Greek Ass story and Apuleius’ Metamorphoses, or Miletus and Sybaris defined the subjects of, respectively, the lost Milesiaca and Sybaritica. To find such a ‘political’ iden-

————— 15 Harrison 1998, 63. 94 GOTTSKÁLK JENSSON tity for the Satyrica we must spend a little more effort than is usual on recon- structing the lost opening episode.

The evidence for the ancient and long independent Greek city of Massalia, or modern Marseilles, as Encolpius’ birthplace comes mainly from one fourth century fragment,16 which, read side by side with a few passages of the Sa- tyrica, yields this information. Servius’ commentary (Aen. 3.57) provides the following description (Fr. I) culled from the full-text Satyrica:

auri sacra fames] sacra id est execrabilis. tractus est autem sermo ex more Gallorum. nam Massilienses quotiens pestilentia laborabant, unus se ex pauperibus offerebat alendus anno integro publicis et purioribus cibis. hic postea ornatus verbenis et vestibus sacris circumducebatur per totam civitatem cum execrationibus, ut in ipsum reciderent mala totius civitatis, et sic proiciebatur. hoc autem in Petronio lectum est.

accursed hunger for gold] sacra means accursed. This manner of speaking derives from the custom of the Gauls, for whenever the Massaliots suffered from a pestilence, one of the poor citizens offered himself to be fed for a whole year on public and pure food. This individual was later dressed in branches and sacred attire and led around the whole city with ————— 16 Fragment IV also ties the Satyrica to Massalia. The fragment consists of a few lines from a poetic eulogy of Sidonius Apollinaris (Carm. 28.145–7): quid vos eloquii canam Latini, / Arpinas, Patavine, Mantuane / … / et te Massiliensium per hortos / sacri stipitis, Arbi- ter, colonum / Hellespontiaco parem Priapo? (‘what shall I sing about you, sires of Latin eloquence, / you from Arpinum, you from Patavium, you from Mantua / ... / and you, Ar- biter, worshipper of the holy tree trunk / throughout the gardens of Massalia, / yourself on par with Hellespontiac Priapus?’). The late fifth century Christian bishop here apos- trophizes three Roman literary worthies (Cicero, Livy and Virgil) by noting only their birthplaces. He goes on to address others and amongst them ‘Arbiter’, who is presented as being in Massalia, as if Petronius Arbiter, the author, were that character of the Sa- tyrica whom Servius refers to in the first fragment. Bücheler 1862, ad Fr. IV, who says the idea had been adumbrated by Lilius Gyraldus, was the first to unravel the biographi- cal fallacy in Sidonius’ reading by noting that the poet ‘thought that Petronius was the same as Encolpius’ (ratus videlicet eundem esse Petronium atque Encolpium). The lines are clearly modeled on Satyrica 139.2, me quoque per terras, per cani Nereos aequor / Hellespontiaci sequitur gravis ira Priapi (‘me, too, through lands, over hoary Nereus’ surface, / haunts the heavy wrath of Hellespontiac Priapus.’) This rather obvious intertex- tuality has, to my knowledge, never been pointed out before, and it shows that no crime against Priapus or resulting wrath of the god is necessitated in Massalia because of fragment IV. THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 95 curses, so that on him would descend the evils of the whole city, and thus he was banished. This can indeed be read in Petronius.

Servius is, unsuccessfully, attempting to explain the word sacra in Virgil by assuming that, since he was a Mantuan and therefore originally from Gallia Cisalpina, he might have used the word in a specifically Gallic sense. Hence the association with the Petronian passage which Servius takes to be reliable evidence for religious customs in the Greek city of Massalia (also in Gallia) in accordance with the grammarian’s practice of culling historical and biographi- cal information from literary texts. It is of scant importance to us whether the information thus acquired is reliable.17 What matters is that Servius read in Petronius that one of the poor citizens of Massalia, unus ex pauperibus [sc. civibus], had volunteered to act the role of ‘scapegoat’ (φαρµακός) in return for being fed for a whole year at public expense, and was then expelled from the city when that time was up. As we learn from textbooks on Greek religion, the human scapegoat is sacrificed only in a social sense. His treatment is reminiscent of that of a beast marked for sacrifice. The beating and cursing of the φαρµακός to ward off sin, plague or famine was no doubt of importance in actual ancient ritual (e.g., the beating of boys in the ritual of Artemis Ortheia at Sparta), but the behavior of the human scapegoat was likely conventionalized and may have resembled that of a writhing dancer or an actor in comedy, such as the stupidus of mime.18 A relevant ancient account of such a ritual survives in the poems of Hip- ponax (Frs. 5–11 [West]). In threatening his enemies with destruction Hippo- nax provides a description of how the φαρµακός should be dealt with: A de- formed and repulsive male is selected and feasted on figs, barley broth, and cheese, then whipped with fig branches and sea onions, and struck seven times on his membrum virile. Walter Burkert explains how there is a moral condem- nation implicit in the rejection of this supposedly depraved individual:19

————— 17 Another commentary, that of Lactantius Placidus on Statius’ Thebais (10.793–4), has sometimes been adduced as further evidence for the historical truth of this alleged Gallic custom, but as Paratore 1933, 1,152, has shown, it is entirely derived from Servius’ clause, using very similar language, and therefore offers no independent evidence. Lac- tantius Placidus mentions neither Massalia nor Petronius. 18 See Wylie 1994, 48–49. 19 For Greek scapegoat rituals generally, see Burkert 1985, 82–84. 96 GOTTSKÁLK JENSSON It is clearly essential that the creature to be driven out be first brought into intimate contact with the community, the city; this is the sense of the gifts of food which are constantly mentioned. Figs are doubly contrasted to normal culture, to the fruits of the field and to the flesh of the victim; they point to sweetness, luxury, licentiousness, a breath of a golden age from which reality must be rudely distinguished …; the outcast is then called the one wiped off all around, peripsema. There is not active kill- ing, but simply a matter of offscourings which must be thrown across the boundaries or over the cliffs, never to return.

It is easy to see how this episode would fit into the Satyrica’s plot. Encolpius, Ascyltos or Eumolpus are just the types to recklessly exploit such a situation without regard for the consequences.20 Constantly penniless and needy, they gladly take every opportunity that comes along to get food, money and sex.21 In the extant Satyrica, religious cults and rituals are generally represented as pretexts for sexual and financial exploitation, and we may accordingly imagine the tone and mood of the episode as anything but solemn.22 But most impor- tantly, the possibility that the branches mentioned in the account of Servius have something to do with the beating of the scapegoat, and, in any case, the general prominence of Encolpius’ phallus in the extant story, make him excep- tionally well suited to play the φαρµακός in such a ritual. In the extant text of the Satyrica the shaving of Encolpius’ and Giton’s heads, which is then inter- preted as sinister for the entire ship’s company (Sat. 103.5), might be cited as a partial parallel. The feeding and fattening of Encolpius is also an important part of the Croton episode,23 where there is likewise a sense of imminent danger which spoils the pleasure of temporary well-being (Sat. 125.2–4). More-

————— 20 Sat. 99.1, ego sic semper et ubique vixi, ut ultimam quamque lucem tamquam non redi- turam consumerem (‘I have always and everywhere lived, as if enjoying the final light and dawn would never rise again’). 21 125.4, nempe rursus fugiendum erit et tandem expugnata paupertas nova mendicitate revocanda (‘no doubt it will be necessary to flee again and our poverty, that had finally been taken care of, will again call for a stint of begging’). 22 Cf. Stephens and Winkler 1995, 365, on the Iolaos and Tinouphis fragments: ‘Both focus on areas in which religion could and often did cross over into charlatanry, both mix in sex and low life in metrical form, both are written in a vigorous but hardly elegant Greek, both are full of textual errors, neither shows a trace of serious purpose.’ 23 125.2, quotidie magis magisque superfluentibus bonis saginatum corpus impleveram (‘each day I filled my stuffed body as the situation with material goods became more and more overabundant’). THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 97 over, his humiliating procession through the streets of Massalia has a partial but striking resemblance to the Risus-festival in Apuleius (Met. 3.1–12), where Lucius is made the butt of the entire citizenry of Hypata.24 Such an opening episode for the Satyrica conforms with the general trend in Greek fiction; a ritual or a religious festival is used in three of the five fully extant erotic fictions to get the plot going (Chariton, Xenophon, Heliodorus). In the Greek cultural context, so preoccupied with the preservation of civic cohesion, to be thus cursed or mocked by a whole city, especially one’s own, is nothing short of a nightmare and certainly the ultimate in humiliation. Ser- vius says he read about this humiliated Massaliot in Petronius, and from the extant part of the Satyrica we know that, of the characters in the story, Encol- pius himself is most susceptible to humiliation. Our reading of the Servius fragment yields information about Encolpius’ citizenship, poverty, voluntary assumption of the degrading role of scapegoat, and final expulsion from Massalia. If this information is right, we would ex- pect some of it to be reflected in what Encolpius says about himself in the fragments of his narrative that have come down to us. Two passages in the extant text of the Satyrica fall into place as soon as we accept this information. Firstly, Encolpius refers to himself as ‘exile’ (81.3, exul) in a retrospective soliloquy at a moment of disillusion when he has no reason to misrepresent himself to the original reader, who already knows the facts; and secondly, Lichas calls him a scapegoat: ‘You thief, what do you have to say for yourself? What stray salamander has burnt off your eyebrows? What god have you offered your hair? Answer, you scapegoat!’ (107.15, ‘quid dicis tu latro? quae sola salamandra supercilia tua exussit? cui deo crinem vovisti? pharmace, responde!’). These retrospective references to the protagonist in the extant Satyrica match so perfectly the fragment of Servius—in both texts Encolpius is an exile and a scapegoat—that their appositeness is most unlikely to be merely coincidental. Let us deal with the latter reference first and then move to the question of exile. That pharmace should be considered Greek, transcribed with Latin letters, is proven by the fact that it occurs nowhere else in extant Latin literature, so far as I have been able to ascertain. As Harlow has shown, pharmace is correctly read as the Greek vocative φαρµακέ, ‘scapegoat’.25 The word belongs to ————— 24 It may be added here in a footnote that Fellini incorporated the Risus-festival into his cinematographic version of the Satyrica, creating some quite memorable scenes. 25 Harlow 1974, 377. 98 GOTTSKÁLK JENSSON the vocabulary of Greek satiric and comic authors such as Hipponax and Aris- tophanes and is used as a term of abuse, and so it might even occur here without a reference to anything specific.26 However, the other two items in the same address do have references to specific facts about Encolpius: He has stolen things from the ship and he has lost his eyebrows. The force of Lichas’ question above is not that he himself believes that a ‘stray salamander’ leaped from the sea aboard the ship and burnt off his eyebrows, but that he is mock- ingly anticipating some far-fetched explanation from Encolpius.27 Rounding off his attack by nastily reminding Encolpius of the humiliation he underwent in Massalia as a ‘scapegoat’, Lichas delivers a serious blow to the ego of our hero. Significantly, Encolpius the narrator acknowledges the truth of Lichas’ accusations: ‘and I couldn’t find anything to say against this accusation of a most obvious guilt’ (108.1, nec quid in re manifestissima dicerem invenie- bam). Let us now examine the description of Encolpius as an exile (81.3) and the significance of this for the story. Besides Encolpius, there are two other ‘exiles’ (exules) in the story. Tryphaena calls Giton an exul (100.4), and she herself is so referred to (100.7, exulem) by Eumolpus, when addressing Encolpius and Giton who would certainly know the facts about her exile.28 It should be noted that the words exilium and exul were not used lightly in the Latin language and rarely in a transferred sense and then only of inanimate things and

————— 26 E.g. Hippon. Fr. 7 [West], et passim; Ar. Ra. 733. The early commentator Janus Souza read pharmace as the vocative of φαρµακός (Burmann 1743, 2:38). LSJ ad verb. derives the abusive sense of φαρµακός, ‘scapegoat’, from the fact that criminals could be used as scapegoats. Strangely, however, translators of the Satyrica have usually taken pharmace here for another Greek word φάρµακος (on the accent see the grammarian Herodianus 1.150 [Lentz]) meaning ‘sorcerer’ and translated it as ‘empoisonneur’ (Ernout), ‘Gift- mischer, Zauberer’ (Stefenelli), ‘poisonous fellow’ (Heseltine), ‘poisonous creature’ (Sullivan), ‘snake in the grass’ (Branham and Kinney). The word is found e.g. in the vocabulary of the Greek LXX. The problem with this reading is that we have no reason to suspect Encolpius of magical practices. 27 A marine animal similar to the salamander, possibly some sort of ‘mollusc’, is said by Pliny (Nat. 10.188) to emit a substance with depilatory effects. 28 Encolpius at one point claims that Ascyltos was ‘by his own admission worthy of exile’ (81.4, sua quoque confessione dignus exilio), which could possibly indicate that we had a fourth exile in the story. The editio Pithoeana has exito, but it is not supported by other witnesses, and shortly before Encolpius has spoken of Ascyltos and himself as having experienced similar fortune (80.8). But even if Ascyltos is an exile he is unlikely to originate from Massalia, because he was clearly not on the ship of Lichas with the others. He is not a protagonist and both comes into the story and disappears from it in Campania. THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 99 animals. Moreover, the terms are without abusive connotations (as opposed to fugitivus, ‘runaway’, ‘fugitive’) since they usually involve people of some rank and standing. An exilium is either a legal banishment (the legal terms are expulsio, eiectio, aquae et ignis interdictio, deportatio and relegatio), or a voluntary emigration (demigratio, fuga, peregrinatio). There is always in these terms an implicit contrast to patria and domus. For these three Greek characters in our story to be called exules in Campania and thereabouts proves that they are not Roman citizens, but come from an independent city outside Ro- man territory. That city is most likely as Greek as they are themselves. The best way to understand the significance of the institution of exile in the Roman world is to consider it in the light of legal arrangements between independent states. An exiled Roman citizen could through the ius exulandi, ‘the right to live in exile’, adopt a new patria and thus forfeit his Roman citizenship.29 This arrangement was reciprocal and exules from independent cities which had a foedus with Rome could take up citizenship there and thus relin- quish their previous status at home (Cic. de Orat. 1.177). In early times the exiled Roman did not need to go far into exile and could find a new home without leaving Latium, in cities such as Tibur, Praeneste, Lavinium and Ardea, or he could go to the Latin colonies. In later times Tarquinii, Nuceria and Ravenna would serve the same purposes. But when the ager Romanus had been expanded so as to cover the whole Italian peninsula and especially after the civil wars, when all Italian cities had been granted Roman citizenship, such places had to be sought outside Italy, in Gallia, Greece or Asia (Cic. Mur. 89). In the early principate the closest foreign city to the North and West, and one that was preeminently qualified to accept Roman exiles, happened to be Mas- salia. This independent Greek city-state, lying in the middle of the Roman province of Gallia Narbonensis, had had a politically privileged status in the area ever since the war against the Gallic tribes in 123–21 B.C.30 In Roman

————— 29 Kleinfeller 1958, 1683–85. 30 Strabo has a chapter on Massalia (Str. 4.1.4f.). In the early principate Massalia was still an officially independent Greek city-state which laid great store by its ancient customs and citizenship and had a long-standing relationship of amicitia with the Romans. In Strabo’s time (ca. 63 B.C. – 19) the city had a high reputation for its rhetorical and philosophical schools which attracted upper class Romans. Tacitus corroborates this reputation of the Massaliots for rhetorical and philosophical skills (Ann. 4.44, Ag. 4). For a concise account of ancient Massalia, see Wackernagel 1966, 2130–53. 100 GOTTSKÁLK JENSSON sources, moreover, it is often mentioned as the preferred destination of Roman exules.31 Given the reciprocity of the institution of exilium, the frequency with which the Romans themselves chose Massalia as their place of exile makes this city the most probable, if not the only possible, place of origin of our first century Greek exules on board a Tarentine ship heading south along the west coast of the Italian peninsula. Since we know that Encolpius is a Massaliot, and we may assume that he left the city by sea on the ship of Lichas, a mer- chant who would have had commercial reasons for going to Massalia, the conclusion is hard to resist that Giton and Tryphaena originate from Massalia, are likewise exiles and were also on that ship. The great complexity of the relationships of Encolpius, Tryphaena, Giton, Lichas and his wife, which is evident from the reciprocal accusations and apologies when the boys board the ship again in the Greek city, requires them to have spent considerable time together on that ship before arriving in Campania. But what could the name of Massalia stand for in Roman and especially Greek literature somewhere around the beginning of our era, when one could imagine that the putative Greek Σατυρικά was written? What would be the

————— 31 Even before the civil war, in 70 B.C., the corrupt former governor of Sicily, C. Verres, chose Massalia as his city of exile and took there much wealth. In 63 B.C. it seemed the obvious place to go to for Catilina, if he had chosen exile (Sal. Cat. 34.2). Milo, too, went there in 52 B.C. (Asc. Mil. 32.13; 45.23), became a citizen and despite his discon- tent could joke that he was happy to be in exile because of the excellent mullets of Mas- salia (D.C. 40.54). After the execution of Jullus Antonius by the order of Augustus, in relation to the adultery of Julia (2 B.C.), his adolescent son Lucius Antonius was sent to Massalia ‘where his exile would be hidden by the pretense of study’ (Tac. Ann. 4.44). Tacitus (Ann. 4.43.5) reports an interesting embassy to Tiberius in 25 undertaken by the Massaliots to ask for the legitimation of the testament of a certain Vulcancius Moschus, who had left his property to the city ut patriae. This well known rhetor (Sen. Suas. 1.2; Con. 2.3.4 et passim) was born in Pergamum (Porphyrion De Hor. ep. 1.4.9), but had to face charges of murder by poison and therefore left Pergamum, despite his defense by Asinius Pollio (Sen. Con. 2.5.13.) and Torquatus, Horace’s friend (Ep. 1.5.9). Later he had settled in Massalia as a rhetor. The Massaliots brought the case before Tiberius to test the validity of the ius exulandi in Massalia, which was thus reaffirmed. Seneca wrote to Nero about a father who had shown his clemency to a son who had made an attempt on his life: ‘satisfying himself with exile—and a luxurious exile—he detained the parri- cide at Massalia and gave him the same liberal allowance that he had before’ (Cl. 1.15.2). Finally, in 58, Nero on false charges bade Cornelius Sulla leave Rome and stay within the walls of Massalia (Tac. Ann. 13.47.3). These walls had been torn down by Caesar in 49 B.C., but were reconstructed by the wealthy Massaliot doctor, Crinas, with Nero’s permission (Plin. Nat. 29.9). THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 101 mythic ‘political’ identity of this city that would lend the story a flavor compa- rable to what we have discussed in the Μιλησιακά and the Μεταµορφώσεις? In Roman literature the name of Massalia is loaded with political and cultural significance. The city’s destiny was perceived as intimately connected with that of Rome from its very foundation. Legend had it that in the times of king Tarquinius the youthful settlers from Phocaea, which is sometimes por- trayed as another sacked Troy (Luc. 3.340), had sailed up the Tiber and made friends with the Romans before continuing on their journey to found Massalia in the midst of savage nations.32 For the Romans they remained a symbol of old Greek civilization miraculously preserved in the heart of barbarian dark- ness.33 Severity, gravity and discipline were the communal virtues of Mas- saliots lauded by Roman authors (Cic. Flac. 26.63; V.Max. 2.6.7). These were virtues that the Romans did not commonly associate with Greeks, but rather with their own ‘old Rome’ (vetus Roma). Massalia was believed to have provided financial aid after the sack of Rome by Gauls, and for this, according to Justin (Just. 43.5.10), it was granted ‘immunity’ (immunitas), ‘an auditors’ place in the senate’ (locus spectaculorum in senatu), and ‘a treaty of equal right’ (foedus aequo iure). Like Rome it fought against the Carthaginians. It had the reputation of a faithful friend and ally to Rome in war and peace (Just. 43.5.3). Accordingly, the siege and subsequent capitulation of Massalia to Caesar during the civil war was perceived as symbolic of the irreparable harm and madness of that conflict. For Rome to turn against such an ally was typical of the self-destructive fraternal slaughter that was the civil war. In the account given by Lucan in the Pharsalia (Luc. 3.298f.) the Massaliots face Caesar with ‘an un-Greek steadfastness’ (Luc. 3.302, non Graia levitate) and they appeal to him by reminding him of the historical relationship of the two states and demonstrating clearly their old-fashioned hatred of tyranny and civil strife. Civil wars are evil, and if Rome has the good fortune to negotiate peace, Cae- sar and Pompey can both come to Massalia to dwell there in exile (Luc. 3.333–5). Thus, in a Roman palimpsest, Massalia, like Troy in the poem of Eu- molpus about the fall of Troy (Sat. 89), might be presented as a projection of Rome herself with respect to her fate in the civil war, the subject of another of

————— 32 There is a short history of Massalia in Justin 43.3–5, which is an epitome of Trogus’ Historiae Philippicae from the first century B.C. 33 This aspect of the city’s image is emphasized in numerous sources: Cic. Flac. 63, Phil. 8.9; Liv. 37.54; Sil. 15.168–72; V.Max. 2.6.7; Tac. Ag. 4; Mela 2.77. 102 GOTTSKÁLK JENSSON Eumolpus’ poems (Sat. 119–124). The logic of the admiration of Massalia by the Romans is explained by A. Trevor Hodge in the following manner: ‘Ro- mans, almost without exception, were fulsome in their admiration, praising the Massaliots as a kind of puritan supermen, while speaking of their politics and foreign policy in terminology that tends painfully to remind a modern ear of a right-winger speaking of a friendly banana republic.’34 Significantly, however, Greek writers have quite a different story to tell, and one that resonates better with the tenor we are familiar with in the story told in our fragments of the Satyrica. They saw Massalia, again in the words of A. Trevor Hodge, ‘as a kind of ‘Naughty Paree, O-la-la!’.’35 In Greek texts the Massaliots have a reputation for being effeminate and soft, which is proven by the fact that they wear floor-length tunics (Athen. Deipn. 12, 523, c; Ps. Plutarch Proverb. Alex. 60). The phrase ‘sailing to Massalia’ (Suidas, s.v. ‘Pleuseias es massalian’; Ps. Plutarch Proverb. Alex. 60) acquired a proverbial meaning of ‘going to the dogs’. Another peculiarity is male proper names with feminine endings: Protis, Apellis, Thespis, Zenothemis, Taxaris, Charmis. The only Roman writer that adopts this Greek attitude towards the Massaliots is the comic dramatist Plautus, who lets a character refer to effeminacy as ‘practising the morals of the Massaliots’ (Plaut. Cas. 963). This atypical attitude for a Roman text, could be explained by Plautus’ own admission that he adapted the Casina from a play by the 3rd cent. Greek poet of New Comedy, Diphilus of Sinope. As far as we can tell, Massalia appears to have been an old-fashioned city state with an aristocratic constitution, and very conservative with respect to its religious customs and the Greek language. An archaic Ionic dialect held its ground there, and Greek continued to be the spoken language until late antiquity.36 Encolpius’ marked preference for old Greek literature and art, and his apparently genuine astonishment at things seen and heard on his trip through Italy under the Julio-Claudian dynasty, are thus intelligible as aspects of his Massaliotic background. His surprise is therefore not due to stupidity, but to his foreign and culturally ‘finer’ origin, and may to a certain extent be used to measure the deviant mores of those whom he encounters. At the same time the ————— 34 Hodge 1998, 4. 35 Hodge 1998, 4. 36 See Clerc 1971, 1: 458–464, on the intellectual culture. The Greek inscriptions of Mas- salia dating from the Roman era are notable for their archaic and Ionian forms, though this may perhaps be due to an officially cultivated archaism to boost local patriotism rather than the survival of the old dialect in common speech. THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 103 joke is always on him, because of his typical Massaliotic effeminacy, his softness and obsession with Giton. There is another reason why this city is especially appropriate as the origin of the narrator of the Satyrica. Massalia, which prided itself on having a port of major commercial importance in the western Mediterranean, was famous for its Atlantic seafarers and their incredible travelogues. Pytheas of Massalia, for one, claimed to have sailed into the outer-sea and north along the coast. His voyage supposedly took him to many previously unknown lands and led to the discovery of the mysterious island of Thule. But he was branded the very worst of liars by Strabo (1.4.3) and mocked by Polybios (34.5.7) as someone too poor—another poor Massaliot—to undertake an expedition to far-away places. Antonius Diogenes certainly parodies Pytheas amongst others in his lost The Wonders beyond Thule (Phot. Bibl. Cod. 166). Euthymenes, another Massaliot adventurer, claimed to have rounded the southern tip of Africa and located the Nile’s source and thus solved this centuries-long debate. But he is called a braggart by the sophist Aelius Aristides and his Periplous nothing but an ‘account for Alkinous’ (Aristid. Aeg. p.354 [Jebb], ἀπόλογος ᾿Αλκίνου), i.e., of the same type as the lying fables told by Odysseus to the gullible king of the Phaeacians. The Younger Seneca cites Euthymenes of Massalia only to refute his claims, and adds that in the olden days ‘there was room for lies; because the realms of the outer sea were unexplored, they were allowed to make up fables’ (Sen. Nat. 4.2.22–25, tunc erat mendacio locus; cum ignota essent externa, licebat illis fabulas mittere). Lucan, with an obvious allusion to Nero’s interest in the problem of the sources of the Nile (Sen. Nat. 6.8.3), also refers to the Massaliot’s story as hearsay, rumor, in a conversation between Caesar and Acoreus, an Egyptian priest (Luc. 10.255–7). Because of such incredible travelers’ tales connected with the city of Massalia, Aelius Aristides uses the term ‘Massaliotic fables’ (Aeg. p.353 [Jebb], µῦθοι Μασσαλιωτικοί) to cover this type of travelogue and relies on his readers to know to what he is referring. Whether the Massaliots Pytheas and Euthymenes were mere liars or misunderstood explorers far ahead of their time makes little difference for our purposes. More important is that they were known to later authors as Odyssean spinners of yarns, which makes their city especially appropriate as the home of Encolpius, the narrator of the travelogue we know as the Satyrica. The outlines of what happened in the first episode in Massalia are not dif- ficult to reconstruct based on such evidence as the Servius fragment, the numerous retrospective allusions in the extant text of the Satyrica and the formu- 104 GOTTSKÁLK JENSSON laic frames of Greek travelogues. Like strangers in Greek literature typically do, Encolpius will have begun his tale by identifying himself through his city of origin, Massalia. Encolpius will further have associated with his Massaliotic identity the qualities that define him most as a character and a narrator: nobil- ity, old-style education and travel, alluding also to his ‘softness’ and love for Giton. His education and taste fit well with the image of Massalia as a univer- sity town in imperial times, and his travels fit well with the fact that Massalia counted among its famous citizens certain travelers who explored the outer- ocean, and came back to tell incredible tales. One does not have to ponder long the possibility of a discursive strategy for the opening of this story to see how the hackneyed motif of the Phaeacian tales of Odysseus could here be given yet another creative spin in Greek literature. The whole set-up is highly adaptable for an ancient Greek satire about literature, human attitudes and morals. The Odyssean traveller who goes from city to city and gets to know many places and the minds of many men is an ideal vehicle for such a satire. Rather than taking a trip to the fabulous edges of the world, as his fellow Mas- saliots claimed to have done, the overeducated but unheroic Encolpius goes to the heart of civilization to face moral and esthetic monstrosities of no less fabulous proportions. This movement inwards to the ordinary (and prosaic) and away from the mythical (and poetic)37 is no doubt related to the therapeu- tic strategy of Greek Cynic satire which ridiculed scholars for studying in de- tail the errors of Odysseus while being ignorant of their own.38 For Petronius

————— 37 There is no room here to discuss the Menippean form of the Satyrica, but very briefly my view is that the genre puts prose in an antagonistic relationship with poetry with a marked preference for prose. This explains a) why the prose is continuous and there is far more of it than the fragmentary poetic passages in the Satyrica; b) why the prose sections are traditionally described as elegant, while the verse has been seen as exceedingly problematic; c) why, in Encolpius’ parlance, speaking in prose is to speak humane, ‘like a human being’, while speaking poetice, ‘like a poet’, is a sure sign of madness (Sat. 90.2– 5); and d) why the poet Eumolpus, a metaliterary figure in a certain sense reflecting the author, is more successful as storyteller in prose, fabulator, than as poeta (as was shown by R. Beck 1979). 38 A valuable but neglected source on the history and ideas of Cynicism is the Cynic letters, most of which derive from the Augustan period (Malherbe 1986, 2 and 14). Diogenes’ Epistles 30–40, in particular, contain material which is often strangely reminiscent of passages in the Satyrica. We have here the same emphasis on the wandering human ex- plorer who goes from city to city and is exemplified by such heroic figures as the beggar Odysseus (34.2–3); we also have striking instances of phallic humour and masturbation (35.2), and perhaps most remarkably the ridicule of stupid signs posted outside private THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 105 the effeminate Massaliot provided, additionally, an ideal vehicle for a satire to subvert Roman chauvinism.

This brings us to the second premise of the argument based on a rereading of the history of the Satyrica’s reception in the late nineteenth and early twentieth century. Habent sua fata libelli. The modern reception begins with Franz Büch- eler’s edition of the text in 1862, which still defines the practices of current editors. In this first modern edition the Satyrica is known as the Saturae or Satires of Petronius, and published in a single volume with the prosimetric satires of Varro and Seneca, along with Priapic poetry. This edition—although the editor identifies the narrative structure and setting as Greek and correctly emphasizes the centrality of the narrative persona of Encolpius—put our text squarely in the class of Roman Satire. Because of the general acceptance of Isaac Casaubon’s classic treatise, De Satyrica Graecorum Poesi & Ro- manorum Satira (Paris 1605), which argued for the necessity of a radical dif- ferentiation between Roman satire and Greek satyric poetry, this meant that any attempt to relate the Saturae of Petronius to Greek texts would provoke accusations of category confusion. Another significant event in the reception of Petronius took place in 1876, when a clever young philologist, Erwin Rohde, published a complete study on the origin of the Greek Romance, Der griechische Roman und seine Vorläufer, which rejected any generic relation between the Satires of Petronius and the extant Greek romances.39 It is well known that Rohde held little appreciation for the Greek Novel, which he saw as being a ‘synthetic’ type of literature: sentimental, because of its origin in erotic poetry; fabulous, because of its origin in fantastic travel literature, or Reisefabulistik; and stylistically preten- tious, because it was written in the Second Sophistic. Although falling outside of his topic, Rohde dedicated one of his extended philological footnotes to the Satires of Petronius, which he identifies as being ‘Menippean satire’. Petronius is following Varro, he concludes, as did Seneca, his contemporary. In laying out the development of Menippean satire, Rohde acknowledges a debt to his friend Friedrich Nietzsche, who had argued that Varro followed Menippus

————— houses in foreign cities (36.1). The best edition is still that of Hercher, Epistolographi Graeci with Latin translation. 39 Rohde 1876, 248–250. 106 GOTTSKÁLK JENSSON closely both with regard to the form and spirit of his satires.40 According to this, if there was any Greek background material to the Satyrica, it had already been filtered through Varro and turned into Roman satire before it could exert influence on Petronius. This anxiety about influence is everywhere present in the classical scholarship of the period, but finds its most pronounced expression in an article published in Hermes two years later, 1878, by Theodor Mommsen, a Berlin pro- fessor towering over a generation of German philologists. The article sets out to accomplish the apparently straightforward task of locating the Campanian city of Trimalchio and analyzing the epigraphic style of the freedman’s pro- jected epitaph, for the purpose of dating the work. However, the impact it had on scholarship derived from a side issue, addressed by the historian with such enthusiasm, and appealing to so many contemporary passions, as to spark a revolution in the study of the Satyrica.41 After praising the account of the adventures of Encolpius and his com- rades as being in the first rank in Roman literature for ‘originality’ and ‘skillful mastery’, Mommsen acknowledges—obliged to do so by Bücheler’s description of the work—that the author of the Satyrica has an obvious fondness for setting the scenes of his story in Hellenic environments, first in Massalia, and then in Greek Campania and Croton. However, despite this fact, Mommsen claims that it is clear that Petronius ‘has, like hardly any other, given full expression to the distinct Italic identity’ (wie kaum ein anderer die italische Indi- vidualität zum vollen Ausdruck gebracht hat), and, ‘perhaps alone of all the Romans, has followed the route of his own genius, independent of Greek models’ (vielleicht allein unter allen römischen unabhängig von griechischen Mustern seinen eigenen genialen Weg gegangen ist). Having formulated this paradox, Mommsen must now offer an explanation of how Petronius could give ‘full expression’ to his ‘Italic identity’ in a work of literature about Greek characters moving in a Greek environment. On the one hand, he argues, Petronius had to be careful not to give any hint of ‘the firm footing of his own nationality’ (den festen Boden der eigenen Natio- nalität) in order not to spoil ‘his setting in an essentially Hellenic environment’ (seine Szene in das eigentlich hellenische Gebiet), but on the other—and

————— 40 Nietzsche 1870, §11 ad fin; in Colli and Montinari 1982, 2: 1, 240f. 41 Mommsen, Th. 1878, 106–121. This article is identified by Bürger 1892, 346 n.2, as the origin of the unprecedented view that the work of Petronius was ‘vielleicht das künstle- risch höchststehende Erzeugniss der ganzen römischen Literatur.’ THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 107 to the same effect—he had no mind to dispense with ‘the influence of the Greek essence’ (die Einwirkungen des griechischen Wesens) in the representation of ‘his home country’ (seiner Heimath) and his times. Mommsen’s Petronius, who is an ‘artist’ (Künstler), and a ‘portrayer of manners and a sati- rist’ (Sittenmaler und Satiriker), was thus constrained to write a Greek story to be faithful to the reality of Hellenization in Italy, and once having embarked on such a project, was forced to conceal his unquestionable ‘Italic nationality’, to which he nevertheless managed to give the fullest expression.42 To understand the modern anxiety at the root of the constitution of Mommsen’s Petronius, we will certainly benefit from paying less attention to the Roman socio-cultural background of the first century, and more attention to the revolutionary events taking place in central Europe in the third quarter of the nineteenth century. The man who invented the modern Petronius was a romantic nationalist and self-confessed animal politicum,43 who had been exiled from Saxony in 1850 for the part he played in the struggle of philologists and other intellectuals for a unified greater Germany. At the time of writing the article on Petronius, Mommsen had been a National Liberal in the Prussian Landtag for five years, and his sympathies towards the recently victo- rious Italian risorgimento movement were obvious and derived from the kin- dred struggle of the two nationalist movements, the German and the Italian, at times against common enemies (e.g. the Garibaldini were greatly aided by Bismarck’s military successes in the Franco-Prussian war). Mommsen’s famous and widely successful Römische Geschichte (1854– 56) was, rather than the history of that ancient empire, the history of the ‘Italic’ nation from the earliest immigrations to the end of the Roman Repub- lic. His interest in ancient history went beyond the scientific, and he con- sciously attempted to write a work of ‘political’ history which would focus on the significance of classical antiquity for his own times. Roman history was his subject of choice, principally because the Italic nation ‘alone among all the civilized nations of antiquity succeeded in constructing a national unity based on political independence’ (errang allein unter allen Kulturvölkern des Alter- tums bei einer auf Selbständigkeit ruhenden Verfassung die nationale Ein-

————— 42 The term ‘Italic nationality’ (italische Nationalität) in Mommsen’s text is meaningless, unless we understand it to be the ancient correlate of the fledgling Italian nationality. Mommsen’s English translator, W.P. Dickson (New York 1868), did not hesitate to translate ‘italische’ with ‘Italian’. 43 In Mommsen’s own testament, Wucher 1956, 218f. 108 GOTTSKÁLK JENSSON heit).44 The terms used by the historian, ‘nation’, ‘independence’, ‘unity’, were the political buzzwords of the time.45 The untenable antithesis in much of Mommsen’s historical writing holds that ‘Roman’ somehow stands for practi- cal realism and a genius for state-building—his Napoleonic Caesar is ‘durch und durch Realist’ (RG. III.450)—while ‘Greek’ is seen as synonymous with fabulous story-telling and abstract philosophizing.46 Although Mommsen does not mention Petronius in his Römische Geschichte, he there molds Terentius Varro into a similar ancient Italian genius. It is Varro’s composition of satires that provided the basis for turning him into a quintessential Roman author, despite ample evidence that he was adapting into Latin a Greek satirical genre. Petronius, likewise, could be recruited as the voice of the Italian nation, because of Bücheler’s use of Satirae as the title of the work, in place of Satyricon or the nominative Satyrica (which as a generic term refers to Greek ‘satyr-plays’), and because this first modern editor of the work had derived it from the satirists Varro and Horace. Even preferring the Latin word to the Greek as the original title of the work is, however, not sufficient to preclude its association with Greek satiric genres, since ancient readers would have directly connected the satires of Petronius with those dis- solute and shameless creatures named σάτυροι.47 The next big step in the modern interpretation of the Satyrica was directly influenced by the arrival on the scene of a new manner of writing, documen- tary Naturalism. As so often in German literature, this movement was heralded by pamphlets demanding a new scientific objectivity in literature. The princi- pal model was Emile Zola and the organs of the movement were such journals as the Kritische Waffengänge (1882–84) in Berlin, and Die Gesellschaft (1885–1902) in Munich. Among the moderately progressive philologists of the day some apprehension was felt that the heritage of classicism and romanti- cism was in danger of being discredited. This concern at least inspired Elimar

————— 44 Römische Geschichte 1854, 1: 30. 45 Wucher 1956, 63. 46 Wucher 1956, 139f. 47 Cf. Schol. Hor. Ep. 1.11.12, saturam … dictam sive a saturis, quod similiter in hoc car- mine ridiculae res pudendaeque dicuntur, quae velut a saturis proferuntur; sive a satura lance; Evanth. de Com. 2.5, satyra … a satyris, quos in iocis semper ac petulantiis deos scimus esse, vocitata est; Schol. Pers. prol. 1, satira … a saturitate, quod plena sit convi- ciis et reprehensionibus hominum; 11.8, dicta … satira a saturitate, unde in choro Liberi patris ministri vino atque epulis pleni Saturi appellabantur. THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 109 Klebs to formulate the first explicit thesis about the composition of the Sa- tyrica, in his classic article, ‘Zur Composition von Petronius’ Satirae’ (1889). Klebs’ Petronius is simply ‘the strongest realist of antiquity’ (der stärkeste Realist des Alterthums) as well as a satirical genius whose great achievement is to have given ‘artistic character’ (künstlerische Charakter) to ‘realism’ (Re- alismus), in contrast with the writers of Klebs’ own time, who ‘merely share with Petronius the long-winded treatment of smut’ (die mit ihm nur die breite Behandlung des Schmutzes gemein haben).48 For Klebs, no attempt is necessary to explain the existence of a realistic novel in antiquity, and so he grants a degree of universality to this predominant form of his times, which enables it to transcend the limitations of literary history. Klebs nevertheless notes the similarities of Encolpius’ narrative persona (an intelligent and well educated person telling the story of his wanderings and chaotic adventures outside the reach of law and civilization) to that of Lucius, in the Greek ass story and the Metamorphoses of Apuleius. But he also finds a partial analogy in the Satyrica with the picaresque novel of the sixteenth and seventeenth centuries. Neither link, however, is seen to have literary-historical implications other than demonstrating the universality of the form. Contrary to what Klebs’ argument has come to represent in the later scholarship,49 his intention was not to argue that Petronius was offering a prosaic parody or travesty of such epic poems as the Odyssey or the Aeneid:

Daran wird natürlich kein Verständiger denken, daß es Petrons Absicht gewesen sei eine prosaische Travestie zu den Gesängen vom Zorn Posei- dons oder Junos zu schreiben. Ein Werk mit einer solchen Fülle lebens- voller Schilderungen der Wirklichkeit erhebt von selber dagegen Ein- spruch, unter die reinen Literatur-Satiren eingereiht zu werden.50

————— 48 A similar definition of the Satyrica as ‘antiker Schmutz’ is found in Nietzsche’s posthu- mous fragments, where the philosopher, after comparing favourably the experience of reading the Satyrica to that of reading the New Testament, poses the following question: ‘ist nicht der antike Schmutz noch mehr werth als diese ganze kleine anmaßliche Chris- ten-Weisheit und -Muckerei?’ Nachgelassene Fragmente; Herbst 1887 bis März 1888 10 [93] (213); in Colli and Montinari 1970, 8: 2, 175–176. 49 E.g. Perry 1967, 186, ‘another sees in it a parody on the epic’, with a footnote reference to Klebs’ article. 50 Klebs 1889, 630. 110 GOTTSKÁLK JENSSON According to Klebs, then, rather than creating a simple parody of epic, Petro- nius merely used an epic structure in the Satyrica for the purpose of achieving ‘inner unity’ (innerer Einheit) for the otherwise loosely structured realistic portrayal of his times. Klebs’ once influential thesis, which postulates an over-arching epic theme of divine wrath in the Satyrica, was in part an expansion of Bücheler’s suggestion that the fragment from Sidonius Apollinaris (Fr. I) might be seen as an indication of Priapic involvement in the story as early as the opening episode in Massalia. To this Klebs added several instances in the extant text where Priapus seems to have a role in the plot. Hence, he concluded that the strife between Encolpius and Priapus was a unifying motif of great importance in the original story. He also drew attention to the many parodic allusions to Greek myth and Roman legends, which serve the same purpose, especially allusions to the Homeric Odyssey, as for instance in the comic recognition scene where Lichas identifies the bald and shaven Encolpius by his mentula, and the narrator explicitly compares this to Odysseus’ more heroic recognition by his scar (Od. 19.386–507). The purpose of such parody in the Satyrica, according to Klebs, is to ex- press, by way of irony, the narrator’s awareness of his pathetic humiliation. This irony is both sophisticated and self-conscious and therefore resembles the narrative posturing frequently assumed by modern authors. To buttress his claim, Klebs highlights the ironic pathos of the narrator where it finds its clearest articulation, in the poem in 139.2, where Encolpius states that the gravis ira Priapi signifies for him what the fateful wrath of Poseidon meant for Odysseus (der Zorn des Priapus bedeutet für Encolpios Schicksale, was Poseidons Zorn für Odysseus). Klebs, in effect, privileges this particular poem and uses it as master text for interpreting the whole of the Satyrica. According to Klebs, by giving the ‘I-novel’ (Ich-Roman) of Encolpius an epic structure, Petronius endowed his Realismus with ‘artistic character’ (künstlerischer Charakter). This supposed achievement of the ancient author is then promoted as the ideal for contemporary writers, an esthetic reconciliation between unre- strained modernity and a possibly endangered classical tradition. The few loose ends that needed tying up—as for example the fact that Petronius wrote the text in the first century, rather than in the nineteenth— were taken care of by the philologist Martin Rosenblüth, whose inaugural dissertation, Beiträge zur Quellenkunde von Petrons Satiren (1902), defined the genre of the Satyrica as ‘synthetic’, i.e. as everything and nothing at the THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 111 same time, and formulated better than others the thesis of the ‘great question mark’ or the generic enigma. We cannot blame German philologists alone for this transformation of the Satyrica, because no protests were heard from Italian, French, English or American scholars. For example, the French scholar Albert Collignon, in his important study Étude sur Pétrone (1892), completely rejected the ‘tempting hypothesis’ (hypothèse séduisante) that there existed a genre of licentious Greek romance which Petronius might have imitated generically. Even the apparently similar Pseudo-Lucianic ass story, he claims, is different, since it is not Menippean in form. In sum, Collignon joins the camp of Mommsen and Rohde, and emphasizes the alleged categorical difference of the Greek and Roman novels: ‘Les romans grecs que nous possédons et le Satiricon ne pro- viennent pas des mêmes sources, et n’ont ni le même objet, ni le même ton’. He also argues with Rohde and Klebs that the Satyrica is a picaresque romance in subject-matter and Menippean satire in form, and he makes of Petronius the inventor of an ‘absolutely original genre’ (une œuvre absolument originale), the Roman novel (roman Latin), the only prototype of which is the Satyrica itself.51 The only classicists who protested were other German philologists, who had witnessed at close hand the alchemy worked upon the ‘antike Schmutz’ of the Satyrica. These scholars were Karl Bürger, but also to a lesser degree Richard Heinze,52 who complained already in the nineteenth century about the easy acceptance of a new scholarly ‘dogma’ that saw the extant Greek and Roman novels as the products of modern-like national sensitivities, and much preferred the Roman, which was seen as original, modern and realistic, to the Greek, which was seen as sentimental and reactionary. It seems that those German scholars who found little of interest for the subject in the exclusive analytical rigour of nationalism, and saw ancient literary history and the connection between Greek and Roman in less black-or- white terms, were mostly ignored by subsequent generations.53 This was un-

————— 51 Collignon 1892, 38–39. 52 Heinze 1899, 494. 53 This view is now changing with a new generation of scholars, as is indicated by the fact that Niklas Holzberg 1995 now follows Heinze completely in his introductory study, and describes the Satyrica throughout in terms of ‘realism’ and parody of the idealistic Greek romance. He further takes the Iolaos fragment as an indication that ‘there really was a Greek tradition of comic realistic narrative combining prose with verse. And it seems reasonable to assume that this tradition was older than Petronius’ Satyrica’ (63). 112 GOTTSKÁLK JENSSON fortunate because the philological arguments still incline towards the case of Latin adaptation from Greek; but, perhaps most compellingly, the verdict of the sands of Egypt, to which both parties to the quarrel had the wisdom to appeal, has been unanimously in favour of Bürger and Heinze and against Mommsen, Rohde and Rosenblüth.54

Later Petronian scholars either tend to brush off the anomaly or fail to notice it at all. It is perhaps a measure of Mommsen’s authority that his attempt to account for the linguistic and cultural mixture of the Satyrica as Petronius’ direct and faithful representation of life in Campania has not been questioned by later scholars. In his seminal study of the ‘Roman novel’, Walsh describes the Satyrica as taking its reader on:

what purports to be a conducted tour of the Greek city-life of Gaul and Italy, but which is essentially a review of the Roman contemporary scene. Though the hero and his friends are Greeks, their attitudes and preoccupations are wholly Roman. The inconsistency did not trouble Petronius, whose aim was ephemeral entertainment, not a closely articu- lated work of art; and the Romanising of the characters and situations lends the novel a greater immediacy and realism [italics mine].55

Mommsen had, to his credit, realized that, in a faithful description of contemporary life, Greek characters moving in a Greek environment should neither speak Latin perfectly like educated Romans, nor quote Roman authors off the top of their heads.56 Walsh merely leaves his readers with a rhetorical antithesis between what ‘purports’ and what ‘is essentially’, which begs the question ————— 54 Rosenblüth’s appeal to the sands of Egypt echoes Bürger’s 1903, 28, final words: ‘Es wäre zu wünschen, daß der Boden Ägyptens, der unsere Kenntnis des idealistischen Ro- mans im Altertume in den letzten Jahren so bedeutend bereichert und uns darüber ganz neue Anschauungen gebracht hat, auch für diesen seinen realistischen Vetter sich einmal fruchtbar erwiese.’ 55 Walsh 1970, 79. 56 The following is a complete list of Greek and Roman authors in the Satyrica: Demosthe- nes (2.5), Homer (2.4, 48.7, 59.3, 118.5), Euripides (2.3), Hyperides (2.8), the nine lyric poets (2.4, 118.5), Pindar (2.4), Plato (2.5), Sophocles (2.3), Thucydides (2.8), Democri- tus (88.3), Eudoxus (88.4), Chrysippus (88.4), Epicurus (104.3, 132.15v.7), Cicero (3.2, 5v.20, 55.5), Lucilius (4.5), Publilius Syrus (55.5), Horace (118.5), Virgil (68.5, 118.5), Cato (137.9v.6), Labeo and Servius (137.9v.8). THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 113 why Petronius should have taken it upon himself to write a Greek story to convey ‘a review of the Roman contemporary scene.’ Walsh’s idea that the very ‘inconsistency’ and artificiality of the ‘Romanising’ of this Greek story could both be entertaining and result in ‘immediacy and realism’ clearly re- calls Mommsen’s topsy-turvy logic. Another scholar who had dealt with the problem just before Walsh, Gareth Schmeling, likewise noted in a study on the personal names in Petronius that:

Greek names so pervade and dominate the Satyricon that the whole at- mosphere becomes Greek … Instead of populating his novel with Greek freedmen, former slaves, and present slaves, Petronius could have used Roman characters. He chose not to. The only literary genre in earlier Roman history to use such a large number of Greek characters was comedy.57

Although Schmeling does not say so, the Greek names in the comedies of Plautus and Terence were taken from the Greek plays that they were adapting into Latin, or invented to suit their Greek context. A likely conclusion, therefore, to be drawn from the similarity of the use of names in the Satyrica and Roman comedy is that the former is a Roman adaptation as well. The point, however, is missed by Schmeling, who claims that Plautus and Terence used Greek names in their plays in order ‘that they might escape the charge of ridiculing and demeaning their own race.’ Schmeling goes on to argue, on the basis of this unfounded Roman chauvinism in authors who were not even true- blooded Romans, that ‘to the Roman audience the use of such a high propor- tion (77%) of Greek names in a work of literature written by a Roman could mean only one thing: comedy’.58 Schmeling’s conclusion is untenable, since Greek names in such texts as Seneca’s adaptations of Greek tragedies were certainly no indication of comedy to their Roman audience. We obviously need to understand better the logic of the linguistic and cultural mixture in the Satyrica. The main character and narrator is a Greek exile from Massalia, who was brought up and educated in the Greek language, but who in the extant part of the work, while a luckless youth wandering in the Greek cities of southern Italy, is represented as fluent in Latin and possessing a mature knowledge of such Roman authors as Cicero, Lucilius, Virgil, Livy ————— 57 Schmeling 1969, 5. 58 Schmeling 1969, 6. 114 GOTTSKÁLK JENSSON and Horace. The native language of Massalia in the early empire was certainly Greek.59 The Massaliot rhetor, Agroitas, whom the Elder Seneca describes as having spoken arte inculta on a certain controversia, in order to resemble a Roman, even so utters his sententia in Greek (Sen. Con. 2.6.12). As a rule, Greek rhetors declaimed in Greek and Roman rhetors in Latin—and possibly Greek, if they had the perfect knowledge of the language that rhetorical exercises demanded. Even if Encolpius is supposed to be a highly atypical Greek who learned Latin as an adult, as for example Dionysius of Halicarnassus claimed to have done,60 which would have made it possible for him to tell his story in Latin, the narrator’s representation of his own youthful self remains problematic. Fresh from Greek speaking Massalia, it is impossible to believe that he would have been so sensitive to the correct pronunciation of Latin that an imperfect recital of Vergilian verses should offend him (Sat. 68.5). In the declamation in Latin which opens the extant text of the Satyrica, this well-trained Greek youth begins by expressing his disgust with bombastic rhetorical exercises, which he describes as filled with fabulous plots and sound-effects, and far removed from the realities of the typically Roman courts in the forum (1.2); he then proceeds to evoke a whole gallery of Greek- only literary worthies (Sophocles, Euripides, Pindar, the lyric poets, Homer, Plato, Demosthenes, Thucydides, Hyperides), who according to him never had to undergo such useless schooling and yet became masters of artistic discourse; and he ends by blaming the decline of oratory on a ‘windy and enor- mous loquacity that has recently migrated to Athens from Asia’ (2.7, nuper ventosa istaec et enormis loquacitas Athenas ex Asia commigravit), referring to the notorious bogeyman of the ‘Asianic’ style, or ‘Asianism’. While it seemed at first that we were situated in the vicinity of the law courts of a Ro- man forum, the bulk of young Encolpius’ inept declamation shows no further awareness of things Roman, but upholds what can only be described as an

————— 59 According to Varro three languages were spoken in Massalia, Greek, Latin and Gallic (Isid. Orig. 15.1.63, Trilingues, quod et graece loquuntur et latine et gallice). But the Gallic language was not written, although undoubtedly spoken by slaves and traders, and Latin was only spoken by the Romans residing in Massalia, at least until the second century, for Latin inscriptions in Massalia are written out in Greek characters (CIL 12.56), and Roman names first begin to appear towards the end of the second century, when Massalia at last became a city under Roman administration; see Clerc 1971, 1: 460. 60 After he settled in Rome where he lived for twenty-two years, as he explains in the introduction to Roman Antiquities. THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 115 Attic point of view, to the extent of having led scholars to suspect that young Encolpius’ language and opinions are ‘owed to a Greek source.’61 This strange mixture of ‘Roman’ and ‘Greek’ is potentially even more confusing in the subsequent Lucilian metrical rendering, improvised by the Greek Agamemnon, on the important subject of the proper schooling for boys. The highly circumlocutory hexameter part of this ‘poem’ could be summa- rized in the following way: Whether born in Athens, Sparta or Naples (sire- numve domus), the boy should begin with Homer, and soon after study Plato and Demosthenes; but then he should switch languages and become immersed in Roman authors and be ‘relieved of the burden of Greek sounds’ (Sat. 5.15– 16, Graio / exonerata sono),62 and when he is thoroughly steeped in Latin literature his taste will change, and he can employ Cicero as model for the composition of epic poetry. An educational programme like this one never existed anywhere in Greco- Roman antiquity. Firstly, there is discrepancy between form and content. Why does Lucilian, and therefore ‘Roman’ satire, deal with the education of Greek schoolboys from Athens, Sparta and Naples? Secondly, the bilingual nature of the curriculum does not square with what we know of the education of Greek boys. Thirdly, it is absurd that the Greek schoolboy would perceive the switch from his own language, Greek, to a foreign language, Latin, as the lifting of a burden. It is true that certain elements here could fit the education of Roman schoolboys, who traditionally began with Greek (the Romans took over wholesale the Greek educational system), before they moved on to works written in Latin. At that point in his education, the Roman boy might well be relieved to switch from a foreign language, Greek, to his own mother tongue, Latin.63 But the poem does not deal with the education of Roman boys. Agamemnon’s school programme is said to be for Greek boys, but it is really for Roman boys, and yet Agamemnon is himself Greek (he does not have a Roman praenomen any more than most of the characters), and he lec- tures in a Greek city (urbs Graeca), where Greek schoolboys would be the norm. Even if we assume, contrary to appearances, that Agamemnon is a thor- ————— 61 See recently Sinclair 1984, 234, who surveys the older scholarship as well. 62 There is another language switcher in the poetic Fr. XXXI. According to Bücheler, Dousa suggested the speaker was a parrot, but even so this parrot would be modeled on the typical advena in Rome. For sonum in the sense of ‘the sound of the spoken language’ see also Ov. Fast. 5.195, corrupta Latino / littera ... Graeca sono. 63 The education of Echion’s son follows the same Roman pattern: 46.5, ceterum iam graeculis [sc. litteris] calcem impingit et Latinas [sc. litteras] coepit non male appetere. 116 GOTTSKÁLK JENSSON oughly Romanized Greek, this highly atypical linguistic condition— improvising poetry in Latin was not an easy feat, even for native speakers— still clashes with the fact that he intends his curriculum for Greek boys.64 The truth is that, however we turn this poem on its head, we can never show that anything of the kind could ever have been composed by any real individual in any real ancient Campanian city. The poem and its setting are simply not, as Mommsen argued, a realistic representation of the cultural mix of southern Italy in the first century. However, if we presume that Petronius recomposed in Latin a preexisting Greek poem on the same topic and shaped it in the form of a Lucilian satire, adding a Roman layer on top of the Greek foundation, this process could well have produced this poem. The underlying Greek hypotext and context would have presented Agamemnon trying to impress Encolpius by improvising in Greek on the topic of how Greek boys had to be raised on the ancient musical diet of Homer (epic), Plato (philosophy), and Demosthenes (rhetoric), so that they could later imitate these canonical authors in their own literary produc- tions. When Petronius reached this poem in his Greek model, in order to re- write it as Lucilian satire, he first had to make changes in the meter. Imitating the most famous contemporary writer of satires in Latin, A. Persius Flaccus (34–62), who imitated the meters of Lucilius in the prologue of his works, the Greek rhetor Agamemnon now breaks into Latin scazons, or limping iambics, and then switches abruptly to hexameters. Towards the end of Petronius’ Latin recomposition, then, the switch of languages is reflected in the boys’ curriculum, and Cicero is added to their reading, regardless of their being as Greek as their teacher. If I have described how Petronius transformed this poem of the Satyrica, then this part at least of his Greek hypotext was just as prosimetric as its Latin adaptation. The unavoidable implication is that the Greek model of the central fabula of the Massaliot Encolpius was prosimetric as a whole. We need not doubt that other sections of the work, such as the shorter fabulae of Eumolpus, both of which are set in Asia Minor, Pergamum and Ephesos, had their Greek models. It is harder to determine, however, whether the large poems attributed to the poet had any counterparts in the Greek model or were just added by Petronius, since the traditional method of Roman adaptation could include ————— 64 The real linguistic constitution of such men was more like that of Lucian’s humiliated Greek scholar in the household of a wealthy Roman pater familias who ‘barbarizes the Roman language’ (Lucian Merc. Cond. 24, τὴν Ῥωµαίων φωνὴν βαρβαρίζων). THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 117 completely new material, or material which came from other works, either Greek or Latin, by ‘contamination’. One amusing side-effect of this thesis is that it seems that we can now finally put to rest the long standing debate about the identity of the city of Trimalchio. In tune with the characteristic layering in the Satyrica of Roman elements on top of Greek foundations, it becomes a possibility that the ‘Greek city’ / ‘Roman colony’ never really existed in ancient Campania, but was created by Petronius through the transformation of the Greek hypotext. Which explains why, despite the fairly detailed description of the place, it has still been impossible to determine its identity to everyone’s satisfaction. Neither the extensive archeological research in the area, nor the great amount of scholarly ink spilled over the problem since Mommsen, has changed much in this respect. The real reason for this state of things is the frustrating inconsistency of the information provided by the text of Satyrica. On the one hand, the place is a ‘Greek city’ with the presence of Greek scholars and a Greek cultural environment (therefore Neapolis), and on the other, the language spoken there is Latin, and it seems that we are dealing with a Roman colony with Roman institutions and magistrates (therefore Puteoli, or even Cumae). Neither Mommsen’s claim that Cumae was properly an urbs Graeca, nor Rose’s con- tention that the term urbs Graeca (Sat. 81.3) is mere mockery of the place—in the manner of Juvenal calling Rome itself a Greek city65—solves the problem. Cumae and Puteoli were not Greek cities by any stretch of the imagination, and the term urbs Graeca issues from the mouth of a native Greek, and is not intended as the mockery of a quintessentially Roman place, but instead refers to a city which shows many signs of being indeed Greek. Perhaps the hardest thing to accept in this new reading of the Satyrica is the idea that the ‘vulgar’ Latin of the freedmen, some of whom are originally of Greco-Asian background, does not represent a realistic imitation of how such characters would actually have spoken Latin. In an interesting twist of the palimpsest, the most fully Romanized Greeks of the Satyrica, such former slaves as Gaius Trimalchio and his friends, speak an inferior Latin compared to the genuine Greeks. The ‘vulgar’ Latin of the freedmen does not betray any unusually strong Greek qualities which would show them to be Petronius’ faithful representation of the speech-mannerisms of this particular ethnic mi- nority in Rome—we have no examples of what the argot of this class was actually like. The modern impression is accidental, and follows directly from ————— 65 Rose 1962, 404; Juv. 3.60–61, non possum ferre, Quirites, / Graecam urbem. 118 GOTTSKÁLK JENSSON Mommsen’s assumption that Petronius set out to document the contemporary scene, in the manner of modern literary Naturalists. Ancient mimicry of speech mannerisms aims at ridiculing the subjects who are imitated, and never shows the modern interest in preserving an accurate image of their ways, for the sake of scientifically inspired documentation. Just as the Latin of the ‘genuine Greeks’ of the story is the generic colloquial Latin of educated Ro- mans (a well defined stylistic category in the rhetorical treatises), the Latin of the Romanized Greeks of the story is the generic language of native speakers among the lower orders, always the legitimate target of ridicule in stratified ancient Mediterranean societies. Although my reading of the Satyrica as palimpsest does deprive the text of a certain imagined quality of ‘presence’ or re-presentation of how things Ro- man really were, it does not propose anything hitherto unheard of with regard to the language of the freedmen. As the best of the seventeenth-century classical scholars understood,66 the freedmen’s language is the literary aping of half- educated colloquial Latin, and as such it is unrelated to the truly uneducated style of the Pompeii graffiti. No doubt the uneducated characters of the Greek work adapted by Petronius spoke a barbaric and solecistic Greek, and Petro- nius decided to retain this feature in his Latin adaptation. The fragments of Greek prosimetric narratives, the Iolaos (POxy. 3010) and Tinouphis (PHaun. inv. 400), show signs of loose writing and ‘vulgarity’ of language.67 Sisenna’s adaptation of the Milesiaca (Μιλησιακά) of Aristides seems to have been in that style too (Fro. Aur. 4.3.2), and judging from the plain language of the epitome of the Greek ass story and Apuleius’ attempts to imply colloquial language without actually writing in that mode, the Metamorphoseis (Μεταµορφώσεις) probably exhibited examples of linguistic mimicry, which in general is a feature of sermocinatio in performa**nce literature. Once the acting of lowly social types hits the stage in performance, the mimicry of their speech mannerisms is irresistible. The nature of the linguistic errors of the freedmen is akin to Trimalchio’s mistakes in mythology; they are errors by design for the sake of humour, since they systematically subvert the correct myths in a way that no true ignoramus

————— 66 E.g., the author of the treatise published in 1666 under the name of Mario Statileo, probably Pierre Petit. The history of the interpretation of ‘vulgar’ Latin in the Satyrica is surveyed in Boyce 1991, 14–34. 67 According to Stephens and Winkler 1995, 367, both texts contain ‘a number of vulgar- isms and uncorrected errors in both the prose and the verse sections of the text.’ THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 119 could accomplish. Trimalchio’s persona is the creation of an educated mind. Niceros’ ghost story (61.3–62.14), likewise, is deliberately mis-told and the character appropriately fears the mocking laughter of the scholastici (Sat. 61.4), not because Latin is his second language, but because he is violating the principles of good rhetorical narration. When all is considered, the language of the freedmen in the Satyrica is no harder to account for in a Roman adaptation of a Greek model than the language of the Greek characters of Plautus, another traditional source for ‘vulgar’ Latin. Trimalchio’s antics at one point offer an interesting example of Latiniza- tion as he overlays the Greek of the Homeric poems with a Latin translation. When his Homeristae are ‘insolently’ exchanging Homeric verses in Greek, he drowns their recital by reading loudly a Latin translation of Homer to his guests (Sat. 59.3). In the same manner of overwriting the Greek voices of the Satyrica, Plocamus, one of Trimalchio’s guests, is made to assert that his own ‘abominable hissing’ is Greek (Sat. 64.5), but the Massaliot Encolpius is un- able to confirm this in his witty Latin narrative, as if his knowledge of Greek was limited to the correct literary Greek of school exercises. A further paradoxical blending is apparent, when the characters Encolpius and Eumolpus, while describing and discussing the works of Greek artists and thinkers, refer to them as ‘Greeks’ (Sat. 83.2, Graeci) and ‘crazy little Greeks’ (Sat. 88.10, Graeculi delirantes), as if they were assuming a patronizing attitude of native Romans towards themselves. Although the latter is obviously ironic, the former is spoken by Encolpius in all seriousness in a simple reference to a Greek term (Sat. 83.2, quam Graeci ‘monocnemon’ appellant). What Encolpius should have said, if he were a simple Latin-speaking Greek, is quam nos ‘monocnemon’ appellamus (‘whom we call “the single-greaved”’). Some- thing strange is going on here, as Müller indirectly admits by wanting to delete ‘Graeci’, just as Fraenkel wanted to delete ‘Graeco more’ in Eumolpus’ description of the type of burial intended in his Ephesian story (Sat. 111.3). Nei- ther place is unsound, however, according to the logic of Roman palimpsests.

To sum up: This paper has argued that Petronius wrote his fictional narrative in the common Roman manner of transforming a pre-existing Greek text. The hypotext, now lost, of his palimpsest most likely bore the same Greek title, Σατυρικά. The argument was constructed in five main stages: Firstly, the literary-historical facts are shown to support the hypothesis of a specific Greek model, i.e. character-names, narrative structure, and testimonia all associate 120 GOTTSKÁLK JENSSON the Satyrica with recognized Roman palimpsests; secondly, a reconstruction of the lost opening episode set in Massalia explains the background and makeup of Encolpius’ personality, the legal status of the protagonists as exiles, and the satirical logic of the Greek story, which, in addition, is shown to be especially suitable for Roman adaptation because of the special significance of Massalia in Latin literature; thirdly, a survey of 19th century scholarship uncovers the anachronistic and ideologically motivated modern reception of the author, Petronius, as a writer of ancient ‘Italian’ realistic fiction; in the fourth place, the cultural mixture of things ‘Roman’ and ‘Greek’ in the Satyrica is shown not to be realistic for Italy in the first century A.D. but, on the contrary, to display utopian qualities characteristic of Roman palimpsests; and in the fifth place, some problematic loci in the text are revisited to show the usefulness of the hypothesis as a tool for improving our understanding of this fragmentary narrative. Finally, I wish to reiterate that it is not an argument against my hypothesis that no such work as a Greek Σατυρικά is ever mentioned in extant literature. We know very well that texts, especially Greek texts of such low repute as semi-pornographic, even ‘criminal’ novels, can easily have disappeared without a trace. And if I am right, this particular work has indeed not disappeared without a trace, because its gist is preserved in the fragments of its Latin palimpsest.

Bibliography

Beck, R. 1979. ‘Eumolpus poeta, Eumolpus fabulator’, Phoenix 33, 239–253. Branham, R B. and Kinney, D. 1997. Petronius: Satyrica, Berkeley – Los Angeles. Bloom, H. 1973. The Anxiety of Influence: A Theory of Poetry, New York. Boyce, B.T. 1991. The Language of the Freedmen in Petronius’ Cena Trimalchionis, Leiden. Bright, D.F. and Ramage, E.S. (eds.) 1984. Classical Texts and Their Traditions: Studies in Honor of C.R. Trahman, Chico, California. Bücheler, F. and Heräus, G. (eds.) 1922 [1862]. Petronii Saturae et Liber Priapeorum … adiectae sunt Varronis et Senecae saturae similesque reliquiae, Berlin. Burkert, W. 1985. Greek Religion, Cambridge, Mass. Burmann, P. (ed.) 1665. Titi Petronii Arbitri Satyricôn Qvae Supersunt, 4 vols. Revised by J.J. Reiske, Amsterdam. Bürger, K. 1888. ‘Zu Apuleius’, Hermes 23, 489–498. — 1892. ‘Der Antike Roman vor Petronius’, Hermes 27, 345–358. — 1902. ‘Der Lukiosroman und seine literaturgeschichtliche Bedeutung’, in Studien zur Geschichte des griechischen Romans. Erster Teil, Blankenburg am Harz. — 1903. ‘Der literaturgeschichtliche Stellung des Antonius Diogenes und der Historia Apollonii’, in Studien zur Geschichte des griechischen Romans. Zweiter Teil, Blanken- burg am Harz. Casaubon, I. 1973 [1605]. De Satyrica Graecorum Poesi & Romanorum Satira, New York. THE SATYRICA OF PETRONIUS AS A ROMAN PALIMPSEST 121 Ciaffi, V. 1955. Struttura del Satyricon, Turin. Clerc, M. 1971 [1927–1929]. Massalia, histoire de Marseille dans l’antiquité. Vol. 1–2, Marseille. Colli, G. and Montinari, M. (eds.) 1970. Nietzsche Werke: Kritische Gesamtausgabe, Berlin. Collignon, A. 1892. Étude sur Pétrone, Paris. Ernout, A. (ed.) 1962 [1923]. Le Satyricon, Paris. Genette, G. 1997 [1982]. Palimpsests: Literature in the Second Degree. (Trans.) Channa Newman and Claude Doubinsky, Lincoln – London. Harlow, R.B. 1974. ‘Pharmace, Petronius 107.15’, Hermes 102, 377. Harrison, S.J. 1998. ‘The Milesian Tales and the Roman Novel’, Groningen Colloquia on the Novel 9, 61–73. Heinze, R. 1899. ‘Petronius und der griechische Roman’, Hermes 34, 494–519. Henriksson, K.-E. 1956. Griechische Büchertitel in der römischen Literatur. Annales Acade- miae Scientiarum Fennicae, Ser. B, Tom. 102,1, Helsinki. Hercher, R. 1873. Epistolographi Graeci, Amsterdam. Heseltine, M. (trans.) 1987. Petronius, Cambridge, Mass. – London. Hinneberg, P. (ed.) 1912. Die griechische und lateinische Literatur und Sprache, in Die Kul- tur der Gegenwart, Teil I, Abtlg. VIII, Berlin. Hodge, A. Trevor. 1998. Ancient Greek France, Redwood, Trowbridge. Holzberg, N. 1993. ‘A Lesser Known ‘Picaresque’ Novel of Greek Origin: The Aesop Ro- mance and its Influence’, Groningen Colloquia on the Novel 5, 1–16. — 1995. The Ancient Novel: An Introduction, London – New York. Jensson, G. 1997. The Recollections of Encolpius: A Reading of the Satyrica as Greco-Roman Erotic Fiction (diss.), Toronto. Klebs, E. 1889. ‘Zur Composition von Petronius Satirae’, Philologus 47, 623–635. Kleinfeller, G. 1958. ‘Exilium’, RE VI.2, 1683–1685. Malherbe, J.A. 1986. [1977]. The Cynic Epistles, Atlanta, Georgia. Mommsen, Th. 1854. Römische Geschichte, Berlin. — 1868. The History of Rome. Translated with the Author’s Sanction and Additions by William P. Dickson; With a Preface by Leonard Schmitz, New York. — 1878. ‘Trimalchios Heimath und Grabschrift’, Hermes 13, 106–121. Nietzsche, F. 1870. Beiträge zur Quellenkunde und Kritik des Laertius Diogenes, Basel. Paratore, E. (ed.) 1933. Il Satyricon di Petronio, Florence. Parsons, P.J. 1971. ‘A Greek Satyricon?’, BICS 8, 53–68. Perry, B.E. 1925. ‘Petronius and the Comic Romance’, CP 20, 31–49. — 1967. The Ancient Romances. A Literary-historical Account of Their Origin, Berkeley – Los Angeles. Rohde, E. 1960 3rd ed. [1876]. Der griechische Roman und seine Vorläufer, Leipzig. — 1893. ‘Zum griechischen Roman’, RhM. 48, 110–140. Rose, C.K.F. 1962. ‘Time and Place in the Satyricon,’ TAPA 93, 402–409. — 1971. The Date and Author of the Satyricon, Leiden. Rosenblüth, M. 1909. Beiträge zur Quellenkunde von Petrons Satiren, Berlin. Schmeling, G.L. 1969. ‘The Literary Use of Names in Petronius’ Satyricon’, Riv. di St. Class. 17, 5–10. Sinclair, B.W. 1984. ‘Encolpius and Asianism (Satyricon 2.7)’, in Bright and Ramage, 231– 237. Smith, M.S. (ed.) 1975. Petronii Arbitri Cena Trimalchionis, Oxford. Stefenelli, A. 1962. Die Volkssprache im Werk des Petron im Hinblick auf die romanischen Sprachen, Vienna. Stephens, S.A. and Winkler, J.J. (eds.) 1995. Ancient Greek Novels: The Fragments, Prince- ton. 122 GOTTSKÁLK JENSSON Sullivan, J.P. 1968. The Satyricon of Petronius: a Literary Study, London. van Rooy, C.A. 1965. Studies in Classical Satire and Related Literary Theory, Leiden. von Wilamowitz-Möllendorff, U. 1907. Die griechische und lateinische Literatur und Spra- che, in P. Hinneberg (ed.) 1912. Teil I, Abtlg. VIII., 2, 124f. Berlin. Wackernagel, H. 1966. ‘Massalia’, RE Band XIV 2, 2130–2153. Walsh, P.G. 1970. The Roman Novel, Cambridge. Williams, G. 1968. Tradition and Originality in Roman Poetry, Oxford. Wucher, A. 1956. Theodor Mommsen. Geschichtsschreibung und Politik, Göttingen. Wylie, K. 1994. Satyric and Heroic Mimes: Attitude as the Way of the Mime in Ritual and Beyond, Jefferson, N.C.